1 CSCE 314: Programming Languages Dr. Flemming Andersen Functional Parsers
What is a Parser? A parser is a program that takes a text (set of tokens) and determines its syntactic structure. String or [Token] Parser syntactic structure 2 3+4 Means Structure is made explicit in our data structure 2
The Parser Type In a functional language such as Haskell, parsers can naturally be viewed as functions. type Parser = String Tree A parser is a function that takes a string and returns some form of tree. However, a parser might not require all of its input string, so we also return any unused input: type Parser = String (Tree,String) A string might be parsable in many ways, including none, so we generalize to a list of results: type Parser = String [(Tree,String)] 3
4 Furthermore, a parser might not always produce a tree, so we generalize the result to a value of any type: type Parser res = String [(res,string)] Finally, a parser might take token streams of symbols instead of character streams: type TokenParser symb res = [symb] [(res,[symb])] Note: For simplicity, we will only consider parsers that either fail and return the empty list of results, or succeed and return a singleton list.
5 Approach and Rationale We are going produce code that will parse complex expressions. The code will be built (like most functional programming) by building from the bottom upwards. This requires a temporary shift in view because we look and think of parsed output from the top down. Let us analyze the following string: To find out how we can parse the syntax and extract its token elements
6 Approach and Rationale We are going produce code that will parse complex expressions. The code will be built (like most functional programming) by building from the bottom upwards. This requires a temporary shift in view because we look and think of parsed output from the top down.
7 Approach and Rationale
protocol ::= http https ftp port ::= 0 1 2 65535 host-domain-name ::= host ::= char char * host. sub-domain. domain sub-domain ::= char *. sub-domain domain ::= com edu gov 8
9 Approach and Rationale We are going produce code that will parse complex expressions. The code will be built (like most functional programming) by building from the bottom upwards. This requires a temporary shift in view because we look and think of parsed output from the top down. We start with very simple parsers When we will glue these together in two ways: One after the other: first an X then a Y afterwards Choice: either an X or a Y. Then this process can be applied recursively. protocol, domain, etc. sub-domain domain 2 25 255 2555
10 Approach and Rationale We are going to produce code that will parse complex expressions. The code will be built (like most functional programming) by building from the bottom upwards. This requires a temporary shift in view because we look and think of parsed output from the top down. High We Level start with Overview very simple parsers When we will glue these together in two ways: We want to be able to write a parser One after the other: first an X then a Y that closely resembles the grammatical structure afterwards the defines the language we re parsing. Choice: The either approach an X or described a Y. Then this process can be applied recursively. next, is involved, but it allows us to do that! protocol, domain, etc. sub-domain domain 2 25 255 2555
11 Basic Parsers (Building Blocks) The parser symboldot consumes the first character and succeeds if it is a dot, and fails otherwise: symboldot :: Parser Char :: String -> [(Char, String)] :: [Char] -> [(Char, [Char])] symboldot (x:xs) (x ==. ) = [(x, xs)] otherwise = [] Example: *Main> symboldot.com [(., com )]
12 Basic Parsers (Building Blocks) The parser item fails if the input is empty, and consumes the first character otherwise: item :: Parser Char :: String -> [(Char, String)] :: [Char] -> [(Char, [Char])] item = \inp -> case inp of [] -> [] (x:xs) -> [(x,xs)] Example: *Main> item "parse this" [('p',"arse this")]
13 The parser return v always succeeds, returning the value v without consuming any input: return :: a -> Parser a return v = \inp -> [(v,inp)] The parser failure always fails: failure :: Parser a failure = \inp -> [] Example: *Main> Main.return 7 "parse this" [(7,"parse this")] *Main> failure "parse this" []
14 We can make it more explicit by letting the function parse apply a parser to a string: parse :: Parser a String [(a,string)] parse p inp = p inp -- essentially id function Example: *Main> parse item "parse this" [('p',"arse this")]
Choice What if we have to backtrack? First try to parse p, then q? The parser p +++ q behaves as the parser p if p succeeds, and as the parser q otherwise. (+++) :: Parser a -> Parser a -> Parser a p +++ q = \inp -> case p inp of [] -> parse q inp [(v,out)] -> [(v,out)] Example: *Main> parse failure "abc" [] *Main> parse (failure +++ item) "abc" [('a',"bc")] 15
16 Examples > parse item "" [] > parse item "abc" [('a',"bc")] > parse failure "abc" [] > parse (return 1) "abc" [(1,"abc")] > parse (item +++ return 'd') "abc" [('a',"bc")] > parse (failure +++ return 'd') "abc" [('d',"abc")]
17 Note: The library file Parsing.hs is available as a link on the course webpage in the lecture table here. The Parser type is a monad, a mathematical structure that has proved useful for modeling many different kinds of computations.
Sequencing Often we want to sequence parsers, e.g., the following grammar: if-stmt ::= if ( expr ) then stmt First parse if, then (, then expr, A sequence of parsers can be combined as a single composite parser using the keyword do. For example: p :: Parser (Char,Char) p = do x item item y item return (x,y) Meaning: The value of x is generated by the item parser. 18
19 Note: Each parser must begin in precisely the same column. That is, the layout rule applies. The values returned by intermediate parsers are discarded by default, but if required can be named using the operator. The value returned by the last parser is the value returned by the sequence as a whole.
20 If any parser in a sequence of parsers fails, then the sequence as a whole fails. For example: > parse p "abcdef" [(( a, c ),"def")] > parse p "ab" [] The do notation is not specific to the Parser type, but can be used with any monadic type.
The Monadic Way Parser sequencing operator (>>=) :: Parser a -> (a -> Parser b) -> Parser b p >>= f = \inp -> case parse p inp of [] -> [] [(v, out)] -> parse (f v) out p >>= f fails if p fails Example otherwise applies f to the result of p this results in a new parser, which is then applied > parse ((failure +++ item) >>= (\_ -> item)) "abc" [('b',"c")] 21
22 Sequencing Typical parser structure p1 >>= \v1 -> p2 >>= \v2 ->... pn >>= \vn -> return (f v1 v2... vn) Using do notation do v1 <- p1 v2 <- p2... vn <- pn return (f v1 v2... vn) If some vi is not needed, vi <- pi can be written as pi, which corresponds to pi >>= \_ ->...
23 Example Typical parser structure rev3 = item >>= \v1 -> item >>= \v2 -> item >>= \_ -> item >>= \v3 -> return $ reverse (v1:v2:v3:[]) Using do notation rev3 = do v1 <- item v2 <- item item v3 <- item return $ reverse (v1:v2:v3:[]) > rev3 abcdef [( dba, ef )] > (rev3 >>= (\_ -> item)) abcde [( e, )] > (rev3 >>= (\_ -> item)) abcd []
24 Key benefit: The result of first parse is available for the subsequent parsers parse (item >>= (\x -> item >>= (\y -> return (y:[x])))) ab [( ba, )]
Derived Primitives Parsing a character that satisfies a predicate: sat :: (Char -> Bool) -> Parser Char sat p = do x <- item if p x then return x else failure Examples > parse (sat (== a )) abc [( a, bc )] > parse (sat (== b )) abc [] > parse (sat islower) abc [( a, bc )] > parse (sat isupper) abc [] 25
Derived Parsers from sat digit, letter, alphanum :: Parser Char digit = sat isdigit letter = sat isalpha alphanum = sat isalphanum lower, upper :: Parser Char lower = sat islower upper = sat isupper char :: Char Parser Char char x = sat (== x) 26
27 To accept a particular string Use sequencing recursively: string :: String -> Parser String string [] = return [] string (x:xs) = do char x string xs return (x:xs) Entire parse fails if any of the recursive calls fail > parse (string "if [") "if (a<b) return;" [] > parse (string "if (") "if (a<b) return;" [("if (","a<b) return;")]
28 many applies the same parser many times many :: Parser a -> Parser [a] many p = many1 p +++ return [] many1 :: Parser a -> Parser [a] many1 p = do v <- p vs <- many p return (v:vs) Examples > parse (many digit) "123ab" [("123","ab")] > parse (many digit) "ab123ab" [("","ab123ab")] > parse (many alphanum) "ab123ab" [("ab123ab","")]
Example We can now define a parser that consumes a list of one or more digits of correct format from a string: p :: Parser String p = do char '[' d digit ds many (do char ',' digit) char ']' return (d:ds) > parse p "[1,2,3,4]" [("1234","")] > parse p "[1,2,3,4" [] Note: More sophisticated parsing libraries can indicate and/or recover from errors in the input string. 29
30 Arithmetic Expressions Consider a simple form of expressions built up from single digits using the operations of addition + and multiplication *, together with parentheses. We also assume that: * and + associate to the right. * has higher priority than +.
Example: Parsing a token space :: Parser () space = many (sat isspace) >> return () token :: Parser a -> Parser a token p = space >> p >>= \v -> space >> return v identifier :: Parser String identifier = token ident ident :: Parser String ident = sat islower >>= \x -> many (sat isalphanum) >>= \xs -> return (x:xs) 31
32 Formally, the syntax of such expressions is defined by the following context free grammar: expr term '+' expr term term factor '*' term factor factor digit '(' expr ')' digit '0' '1' '9'
33 However, for reasons of efficiency, it is important to factorize the rules for expr and term: expr term ('+' expr ε) term factor ('*' term ε) Note: The symbol ε denotes the empty string.
34 It is now easy to translate the grammar into a parser that evaluates expressions, by simply rewriting the grammar rules using the parsing primitives. That is, we have: expr :: Parser Int expr = do t term do char '+' e expr return (t + e) +++ return t expr term ('+' expr ε) term factor ('*' term ε)
35 term :: Parser Int term = do f factor do char '*' t term return (f * t) +++ return f expr term ('+' expr ε) term factor ('*' term ε) factor :: Parser Int factor = do d digit return (digittoint d) +++ do char '(' e expr char ')' return e
36 Finally, if we define eval :: String Int eval xs = fst (head (parse expr xs)) then we try out some examples: > eval "2*3+4" 10 > eval "2*(3+4)" 14 > eval "2+5-" 7 > eval "+5-" *** Exception: Prelude.head: empty list